1. Field of the Invention
The present invention relates generally to an apparatus and method for precoding in a wireless communication system, and in particular, to an apparatus and method for precoding by limited cooperation in a wireless communication system.
2. Description of the Related Art
In a multi-cell environment, a User Equipment (UE) located at a cell boundary simultaneously receives signals from its own cell, i.e., a serving cell, and from an adjacent cell. The signal received from the adjacent cell acts as an interference with the UE, thus degrading the performance of the UE.
Therefore, research is conducted on various precoding schemes for efficiently controlling such interference. Typical examples of the precoding schemes include a Maximum Ratio Transmission (MRT) scheme, an Interference Minimizing Transmission (IMT) scheme, and a Signal-to-Leakage-plus-Noise Ratio (SLNR) maximization scheme.
The MRT scheme transmits signals by forming beams according to the channel of a UE supported by each cell. The MRT scheme maximizes a reception (RX) Signal-to-Noise Ratio (SNR) of a UE supported by each cell, without considering possible interference with another cell. However, because each cell transmits signals by considering only the relationship with a UE supported by the cell, without considering the communication environment of another cell, the MRT scheme cannot control interference from an adjacent cell to a UE located in another cell with a poor environment, thus degrading the performance of the UE.
The IMT scheme transmits signals while minimizing possible interference from each cell to another cell. In the IMT scheme, each cell collects channel information about all UEs supported by the cell, generates an aggregate interfered channel matrix, and uses a singular vector corresponding to the minimum singular value as a precoding vector through singular value decomposition, thereby minimizing a possible interference with another UE. In this manner, the IMT scheme controls interference with another cell, rather than providing smooth communication in each cell. Therefore, a UE located in another cell with a poor environment has a reduced interference from an adjacent cell, thus having an improved performance. However, a UE located in a cell with a good environment cannot receive more data due to a precoding vector set to reduce interference (can otherwise receive more data), thus having a degraded performance.
The SLNR maximization scheme is a favorable combination of the MRT scheme and the IMT scheme. The SLNR maximization scheme transmits signals by forming beams according to the channel of a UE supported by each cell, while minimizing al possible interference from each cell to another cell. Herein, the SLNR is a ratio of the size of a transmission (TX) signal to a sum of the size of interference (leakage) from the TX signal to another cell and the size of a noise. That is, the SLNR maximization scheme maximizes an SLNR to secure the quality of a TX signal and the performance of a UE in another cell. However, the SLNR maximization scheme fails to reflect the practical communication environments of other cells because each cell sets a precoding vector by considering only the channel of possible interference with another cell.
For example, assuming that there are three cells, where first, second, and third Base Stations (BSs) transmit signals to respective first, second, and third UEs, a channel between the first BS and the second UE and a channel between the first BS and the third UE are in the same state, and the communication environment of the third BS is poorer than the communication environment of the second BS, because the first BS causes large interference to the second UE, the first BS sets a precoding vector by considering the second UE more than the third UE.
However, in this example, the second BS is not greatly affected by interference because of its good communication environment, while the third BS is greatly affected by even a small interference because of its poor communication environment. Thus, the precoding vector set by the first BS fails to effectively control interference with another cell. Also, because the conventional SLNR maximization scheme considers only two cells, it is an inefficient interference control scheme in an environment where there are three or more cells.